New issue
Advanced search Search tips
Note: Color blocks (like or ) mean that a user may not be available. Tooltip shows the reason.
Starred by 2 users

Issue metadata

Status: WontFix
Owner: ----
Closed: Oct 2010
Cc:



Sign in to add a comment

Localized system paths are not being found on windows XP

Reported by christop...@gmail.com, Dec 9 2008

Issue description

german WinXP SP3, Python 2.5.2, DELL Inspiron 9400 laptop with 2GB Ram, 250 
GB Harddrive

On a german system, the default plugin system directory for the firefox 
plugin is not found, as the build script is assuming a wrong path.

What steps will reproduce the problem?
1. Windows XP German
2. walk through the "Getting Started" guide to build the firefox plugin
3. the compilation failes, as the build script expects to build into 
"C:\Program Files\Mozilla Firefox\", whereas on localized versions of 
windows, the program directory can have different names. E.g. in german, it 
is "C:\Programme"

I see this when doing the steps in the "Getting started" guide:

C:\nacl\googleclient\native_client>.\scons.bat --prebuilt firefox_install
scons: Reading SConscript files ...
======================================================================
Building nexe binaries using sdk at 
[C:\nacl\googleclient\native_client\tools_bi
n\windows\sdk\nacl-sdk]
======================================================================
scons: done reading SConscript files.
scons: Building targets ...
C:\Python25\python.exe ./tools/firefoxinstall.py MODE=0 
PLATFORM_BASE="C:\nacl\g
oogleclient\native_client/scons-out/"
MODE=0
PLATFORM_BASE=C:\nacl\googleclient\native_client/scons-out/
Error: directory "c:\Program Files\Mozilla Firefox\plugins" not found.
I don't know where to install browser plugins.

**** INSTALLATION FAILED.

scons: unknown OSError exception code 0 - 'C:\WINDOWS\System32\cmd.exe /C 
""C:\P
ython25\python.exe" ./tools/firefoxinstall.py MODE=0 
PLATFORM_BASE="C:\nacl\goog
leclient\native_client/scons-out/""': Error
scons: *** [firefox_install] Error 127
scons: building terminated because of errors.


Workaround is to manually create the english directory manually 
("C:\Program Files\") and then copy the Mozilla firefox folder into it, and 
then recopy the plugins folder back to the normal directory.

As an alternative, the build script could also create non-existing 
directories on the fly in order to create the output, but this would still 
not make the plugin work. 

Best would be to find the localized application folder or to let the user 
specify where it is if it is not found in the assumed location. 


 
"c:\Program Files\Mozilla Firefox\plugins" is hard coded somewhere: this is not
related to localization. The installation failed on my system: the files were copied
to "c:\Program Files\Mozilla Firefox\plugins" and my windows drive is J:.


Labels: -Type-Defect Type-Enhancement
Status: Started
Henry - if you don't object I'm going to suggest we don't fix this. Developers will
be obliged to use an english machine for the forseeable future.
Speaking as someone living in a country with two official languages (Canada - English
and French) I think forcing English developer support is a *bad* idea.

Hard-coded paths make everyone's life more difficult. You lose developers who should
be testing and enhancing, and alienate people. Think of all those great Estonian
hackers and cryptographers (no, I'm not kidding, there are some great people there)
who won't look at your source because their OS has spoken Estonian for years...

Also, you'll be finding those hard-coded English paths FOREVER. Been there and done
that; I speak from experience.

Rather than full i18n support, a small step might be to just have one config file
where all those pesky "C:\Windows" things can be configured. Just my $0.02 (CAD).
Hi....

Have you try to find a firefoxinstall file? This configuration is hard code... 

Follow this steps:

- Ctrl + F for a "def Windows_install" method;
- In this method, looking for a call "os.path.join";
- Change de "Program File" parameter to your directory.

P/s. I have the same problem... My language windows installation is Portuguese 
(Brazil)... And I change to my idiom and now the NACL is work fine...

I hope that post helps you
We ran into similar issues during development of a .NET application.  This issue
exists in many localized versions of Windows XP (Italian is "Programmi" for example).
Our solution was to use the Windows dynamic links:
%programfiles%
%temp%
etc.
This should find the system's default "Program Files" path on any version of Windows.
Here's an article about using these Environment Variables.  The Windows default PATH
variables are detailed near the end.
http://en.wikipedia.org/wiki/Environment_variable

Labels: -Priority-Medium Priority-Low HelpWanted
Status: Available
The team is more focused on Chromium integration than plugin support right now.  
Demoting and labeling as HelpWanted.  Patches welcome.
Labels: OpSys-Windows
Project Member

Comment 10 by bradnelson@google.com, Dec 3 2009

Comment 11 by hbridge@google.com, Jan 29 2010

Labels: -Priority-Low Pri-3

Comment 12 by hbridge@google.com, Jan 29 2010

Labels: -OpSys-Windows OS-Windows

Comment 13 by hbridge@google.com, Oct 15 2010

Status: WontFix
Project Member

Comment 14 by bugdroid1@chromium.org, Jan 21 2011

The following revision refers to this bug:
    http://src.chromium.org/viewvc/chrome?view=rev&revision=1958

------------------------------------------------------------------------
r1958 | adonovan@google.com | Mon Apr 05 14:40:53 PDT 2010

Changed paths:
 M http://src.chromium.org/viewvc/chrome/trunk/src/native_client/SConstruct?r1=1958&r2=1957&pathrev=1958
 M http://src.chromium.org/viewvc/chrome/trunk/src/native_client/tools/llvm/toolchain-creator.sh?r1=1958&r2=1957&pathrev=1958
 M http://src.chromium.org/viewvc/chrome/trunk/src/native_client/tools/llvm/llvm-fake.py?r1=1958&r2=1957&pathrev=1958

Various fixes:
- Issue a warning if naclsdk_mode=manual is specified with ARM.
- Remove naclsdk_mode=manual from toolchain-creator.sh.
- Switch LIBDIR_X8632_{1,2}
- Read LLC_SFI_X86 from environment.
- Don't splice in the REACHABLE_FUNCTION_SYMBOLS in ld.

Review URL: http://codereview.chromium.org/1613004
------------------------------------------------------------------------
Project Member

Comment 15 by bugdroid1@chromium.org, Mar 18 2011

The following revision refers to this bug:
    http://src.chromium.org/viewvc/chrome?view=rev&revision=4567

------------------------------------------------------------------------
r4567 | robertm@google.com | Fri Mar 18 08:31:47 PDT 2011

Changed paths:
 M http://src.chromium.org/viewvc/chrome/trunk/src/native_client/tests/spec2k/run_all.sh?r1=4567&r2=4566&pathrev=4567
 M http://src.chromium.org/viewvc/chrome/trunk/src/native_client/buildbot/buildbot_selector.py?r1=4567&r2=4566&pathrev=4567
 A http://src.chromium.org/viewvc/chrome/trunk/src/native_client/buildbot/buildbot_pnacl2.sh?r1=4567&r2=4566&pathrev=4567

Added new config for pnacl2 bot to test spec 2k in train mode for all platforms.

BUG= http://code.google.com/p/nativeclient/issues/detail?id=1497
TEST= gcl try bots --bot nacl-lucid64-pnacl2
Review URL: http://codereview.chromium.org/6665045
------------------------------------------------------------------------
Project Member

Comment 16 by bugdroid1@chromium.org, Mar 29 2011

The following revision refers to this bug:
    http://src.chromium.org/viewvc/chrome?view=rev&revision=4651

------------------------------------------------------------------------
r4651 | jvoung@google.com | Mon Mar 28 17:04:38 PDT 2011

Changed paths:
 M http://src.chromium.org/viewvc/chrome/trunk/src/native_client/tests/spec2k/176.gcc/run.train.sh?r1=4651&r2=4650&pathrev=4651

Narrow 176.gcc validation exclusion to  issue 893  (x86-32 pnacl).

TEST=
- confirm that x86-32 is still an issue.
- run spec on all other platforms to confirm it is not an issue on those.
- gcl try THISCL --bot nacl-lucid64-pnacl2
  (to double check)

BUG= http://code.google.com/p/nativeclient/issues/detail?id=893
Review URL: http://codereview.chromium.org/6726038
------------------------------------------------------------------------
Project Member

Comment 17 by bugdroid1@chromium.org, Apr 19 2011

The following revision refers to this bug:
    http://src.chromium.org/viewvc/chrome?view=rev&revision=4930

------------------------------------------------------------------------
r4930 | robertm@google.com | Tue Apr 19 09:55:38 PDT 2011

Changed paths:
 M http://src.chromium.org/viewvc/chrome/trunk/src/native_client/buildbot/buildbot_selector.py?r1=4930&r2=4929&pathrev=4930
 M http://src.chromium.org/viewvc/chrome/trunk/src/native_client/buildbot/buildbot_pnacl1.sh?r1=4930&r2=4929&pathrev=4930

Activate buildbot, lucid64-pnacl2, for pnacl.
Before it was just running a nacl-gcc tests.
Also, change task assignement as follow to reduce bot running time:
 lucid64-pnacl1   x86-32 only
 lucid64-pnacl2   x86-64 only

BUG= http://code.google.com/p/nativeclient/issues/detail?id=1647
Review URL: http://codereview.chromium.org/6883048
------------------------------------------------------------------------
Project Member

Comment 18 by bugdroid1@chromium.org, Apr 28 2011

The following revision refers to this bug:
    http://src.chromium.org/viewvc/chrome?view=rev&revision=5089

------------------------------------------------------------------------
r5089 | robertm@google.com | Thu Apr 28 13:26:56 PDT 2011

Changed paths:
 M http://src.chromium.org/viewvc/chrome/trunk/src/native_client/buildbot/buildbot_selector.py?r1=5089&r2=5088&pathrev=5089
 M http://src.chromium.org/viewvc/chrome/trunk/src/native_client/buildbot/buildbot_pnacl2.sh?r1=5089&r2=5088&pathrev=5089

Proposal for reorganzation of spec build-bots.

Trying to get some early feedback, this will ultimately get rid of "tests/spec2k/bot_spec.sh".

We also would need one additional build bot, and some bot renaming.

TEST= buildbot/buildbot_pnacl2.sh mode-spec-pnacl-arm
      buildbot/buildbot_pnacl2.sh mode-spec-pnacl-trybot

Review URL: http://codereview.chromium.org/6883166
------------------------------------------------------------------------
Project Member

Comment 19 by bugdroid1@chromium.org, May 17 2011

The following revision refers to this bug:
    http://src.chromium.org/viewvc/chrome?view=rev&revision=5333

------------------------------------------------------------------------
r5333 | pasko@google.com | Tue May 17 03:51:10 PDT 2011

Changed paths:
 M http://src.chromium.org/viewvc/chrome/trunk/src/native_client/DEPS?r1=5333&r2=5332&pathrev=5333

Update toolchain DEPS.

Bad news:

Updating arm_toolchain_version from 5262 to 5330 reveals trybot failure on nacl-lucid64-pnacl2:
http://chromegw.corp.google.com/i/tryserver.nacl/builders/nacl-lucid64-pnacl2/builds/1311

stage: spec2k build [SetupPnaclArmOpt] [train-some] [spec2k build [SetupPnaclArmOpt] [train-some]]
stderr        : gcc.opt.bc: error: undefined reference to 'signal'
gcc.opt.bc: error: undefined reference to 'system'

this passed before the roll.
*Not* rolling the ARM toolchain revision now.
Review URL: http://codereview.chromium.org/7017009
------------------------------------------------------------------------
Project Member

Comment 20 by bugdroid1@chromium.org, May 20 2011

The following revision refers to this bug:
    http://src.chromium.org/viewvc/chrome?view=rev&revision=5382

------------------------------------------------------------------------
r5382 | robertm@google.com | Fri May 20 11:25:33 PDT 2011

Changed paths:
 M http://src.chromium.org/viewvc/chrome/trunk/src/native_client/tools/llvm/utman.sh?r1=5382&r2=5381&pathrev=5382
 A http://src.chromium.org/viewvc/chrome/trunk/src/native_client/tools/llvm/utman-test.sh?r1=5382&r2=5381&pathrev=5382 (from /trunk/src/native_client/tools/llvm/utman.sh revision 5379)

This CL splits out the testing subpart of utman.sh into a separate testing 
script: utman-test.sh

There is essentially no coupling between testing and the rest of the utman 
script  so this will reduce the size of the massive utman.sh by 10% without 
much surgery.

Going forward this separation allows us to unify utman-test.sh
with  the build bot scripts
buildbot/buildbot_pnacl1.sh and buildbot/buildbot_pnacl2.sh.
This should help eliminate code duplication and ensure that we run
the same tests for local pnacl toolchain builds and on the bots.

BUG= http://code.google.com/p/nativeclient/issues/detail?id=1647
Review URL: http://codereview.chromium.org/7004033
------------------------------------------------------------------------
Project Member

Comment 21 by bugdroid1@chromium.org, Jul 6 2011

The following revision refers to this bug:
    http://src.chromium.org/viewvc/chrome?view=rev&revision=5911

------------------------------------------------------------------------
r5911 | robertm@google.com | Wed Jul 06 10:00:40 PDT 2011

Changed paths:
 D http://src.chromium.org/viewvc/chrome/trunk/src/native_client/tests/spec2k/bot_spec.sh?r1=5911&r2=5910&pathrev=5911

Removed an unused script which has been superseded by buildbot/buildbot_pnacl2.sh
Review URL: http://codereview.chromium.org/7057016
------------------------------------------------------------------------
Project Member

Comment 22 by bugdroid1@chromium.org, Jul 19 2011

The following revision refers to this bug:
    http://src.chromium.org/viewvc/chrome?view=rev&revision=6111

------------------------------------------------------------------------
r6111 | jvoung@google.com | Mon Jul 18 17:19:02 PDT 2011

Changed paths:
 M http://src.chromium.org/viewvc/chrome/trunk/src/native_client/tests/pnacl_client_translator/nacl.scons?r1=6111&r2=6110&pathrev=6111
 M http://src.chromium.org/viewvc/chrome/trunk/src/native_client/DEPS?r1=6111&r2=6110&pathrev=6111

Bump toolchain DEPs to pull in pnacl-ld fix, and turn on
pnacl_srpc_hw{1,2}_browser_test, regardless of platform.

TEST= trybots, run_pnacl_srpc_hw2_browser_test w/ platform=x86-64
Review URL: http://codereview.chromium.org/7411005
------------------------------------------------------------------------
Project Member

Comment 23 by bugdroid1@chromium.org, Jul 29 2011

The following revision refers to this bug:
    http://src.chromium.org/viewvc/chrome?view=rev&revision=6239

------------------------------------------------------------------------
r6239 | pdox@google.com | Thu Jul 28 18:30:52 PDT 2011

Changed paths:
 A http://src.chromium.org/viewvc/chrome/trunk/src/native_client/tools/llvm/hgrc?r1=6239&r2=6238&pathrev=6239
 M http://src.chromium.org/viewvc/chrome/trunk/src/native_client/tools/llvm/merge-tool.sh?r1=6239&r2=6238&pathrev=6239
 M http://src.chromium.org/viewvc/chrome/trunk/src/native_client/buildbot/buildbot_selector.py?r1=6239&r2=6238&pathrev=6239
 M http://src.chromium.org/viewvc/chrome/trunk/src/native_client/tools/llvm/utman.sh?r1=6239&r2=6238&pathrev=6239
 D http://src.chromium.org/viewvc/chrome/trunk/src/native_client/buildbot/buildbot_pnacl2.sh?r1=6239&r2=6238&pathrev=6239
 A http://src.chromium.org/viewvc/chrome/trunk/src/native_client/buildbot/buildbot_spec2k.sh?r1=6239&r2=6238&pathrev=6239 (from /trunk/src/native_client/buildbot/buildbot_pnacl2.sh revision 6231)
 A http://src.chromium.org/viewvc/chrome/trunk/src/native_client/buildbot/buildbot_pnacl_merge.sh?r1=6239&r2=6238&pathrev=6239

1) Add a new mode to utman for downloading "upstream" repository, which contains both LLVM and LLVM-GCC.

2) Rename buildbot_pnacl2.sh to buildbot_spec2k.sh, along with some cleanup.

2) Setup two new builders "pnacl-merge-fast" and "pnacl-merge-slow" for testing LLVM merging.

pnacl-merge-fast:
  Perform merge, rebuild PNaCl, run a series of quick tests.
  If all the quick tests pass, commit the merge.
  This bot is not yet enabled, pending a merge to tip.

pnacl-merge-slow:
  Using the last committed merge result, run all tests, including browser tests and Spec2K.
  This keeps track of the total status of the merge.
Review URL: http://codereview.chromium.org/7524023
------------------------------------------------------------------------
Project Member

Comment 24 by bugdroid1@chromium.org, Oct 31 2011

The following revision refers to this bug:
    http://src.chromium.org/viewvc/native_client?view=rev&revision=7061

------------------------------------------------------------------------
r7061 | robertm@google.com | Mon Oct 31 14:59:47 PDT 2011

Changed paths:
 M http://src.chromium.org/viewvc/native_client/trunk/src/native_client/buildbot/buildbot_selector.py?r1=7061&r2=7060&pathrev=7061
 M http://src.chromium.org/viewvc/native_client/trunk/src/native_client/buildbot/buildbot_pnacl.sh?r1=7061&r2=7060&pathrev=7061

Cleanup and simplify build bot scripts for pnacl.

Brad: 
we still show the two dead trybots:

nacl-lucid64-pnacl1	
nacl-lucid64-pnacl2


on the waterfall http://chromegw.corp.google.com/i/tryserver.nacl/waterfall
how do we get rid of the titles?
Review URL: http://codereview.chromium.org/8433001
------------------------------------------------------------------------
Project Member

Comment 25 by bugdroid1@chromium.org, Dec 11 2011

The following revision refers to this bug:
    http://src.chromium.org/viewvc/native_client?view=rev&revision=7408

------------------------------------------------------------------------
r7408 | pdox@google.com | Sat Dec 10 19:23:17 PST 2011

Changed paths:
 M http://src.chromium.org/viewvc/native_client/trunk/src/native_client/SConstruct?r1=7408&r2=7407&pathrev=7408

Add -Wno-array-bounds to SConstruct for PNaCl GlibC build.

GlibC's strpbrk is implemented by a macro which uses short-circuiting with
the ternary operator to quickly evaluate the case for 0, 1, 2, or 3 length strings.
For some reason, this construction triggers Clang's array-bounds detection.

I've filed an LLVM bug for this:
http://llvm.org/bugs/show_bug.cgi?id=11536

TBR= jvoung
BUG= http://llvm.org/bugs/show_bug.cgi?id=11536
TEST= PNaCl GLibC bot
Review URL: http://codereview.chromium.org/8911001
------------------------------------------------------------------------
Project Member

Comment 26 by bugdroid1@chromium.org, Jul 17 2012

The following revision refers to this bug:
    http://src.chromium.org/viewvc/chrome?view=rev&revision=147088

------------------------------------------------------------------------
r147088 | yfriedman@chromium.org | Tue Jul 17 13:45:48 PDT 2012

Changed paths:
 M http://src.chromium.org/viewvc/chrome/trunk/src/build/common.gypi?r1=147088&r2=147087&pathrev=147088

[Android] Additional fixes to build/common.gypi for Android.

Fixes:
1) disable_nacl
2) disable p2p_apis
3) Enable breakpad for official builds
4) Alpha-short use_system_icu
5) Enable .relro sections
6) Remove defunct android_unit_test_target_type flag

BUG=137569


Review URL: https://chromiumcodereview.appspot.com/10783026
------------------------------------------------------------------------
Project Member

Comment 27 by bugdroid1@chromium.org, Jan 17 2014

The following revision refers to this bug:
    http://src.chromium.org/viewvc/native_client?view=rev&revision=12635

------------------------------------------------------------------------
r12635 | mcgrathr@chromium.org | 2014-01-17T23:35:21.538805Z

Changed paths:
   M http://src.chromium.org/viewvc/native_client/trunk/src/native_client/tools/REVISIONS?r1=12635&r2=12634&pathrev=12635
   M http://src.chromium.org/viewvc/native_client/trunk/src/native_client/toolchain_build/toolchain_build.py?r1=12635&r2=12634&pathrev=12635
   M http://src.chromium.org/viewvc/native_client/trunk/src/native_client/tools/Makefile?r1=12635&r2=12634&pathrev=12635

Don't pass -D_I386MACH_ALLOW_HW_INTERRUPTS explicitly in newlib build

This brings in recent newlib changes:

cc9ce45 Add -D_I386MACH_ALLOW_HW_INTERRUPTS in configure.host for nacl
489f1d8 Include more of sys/types.h and sys/signal.h under NaCl
02b55a2 Define _POSIX_THREADS under NaCl.

The most recent change makes it superfluous to pass the -D switch
via configure options, so stop doing that.

BUG= cleanup
TEST= toolchain trybots
R=bradnelson@google.com

Review URL: https://codereview.chromium.org/137843003
------------------------------------------------------------------------
Project Member

Comment 28 by bugdroid1@chromium.org, Feb 4 2014

The following revision refers to this bug:
    http://src.chromium.org/viewvc/native_client?view=rev&revision=12686

------------------------------------------------------------------------
r12686 | sbc@chromium.org | 2014-02-01T22:53:25.018616Z

Changed paths:
   M http://src.chromium.org/viewvc/native_client/trunk/src/native_client/pnacl/COMPONENT_REVISIONS?r1=12686&r2=12685&pathrev=12686
   M http://src.chromium.org/viewvc/native_client/trunk/src/native_client/tools/REVISIONS?r1=12686&r2=12685&pathrev=12686
   M http://src.chromium.org/viewvc/native_client/trunk/src/native_client/toolchain_build/toolchain_build.py?r1=12686&r2=12685&pathrev=12686

Update newlib version in tools/REVISIONS

This brings in the following newlib change:
41e7008: Revert "Fix race condition in concurrent printf and fopen."

For the pnacl toolchain which was lagging it also brings
in the following four changes:
cc9ce45: Add -D_I386MACH_ALLOW_HW_INTERRUPTS in configure.host for nacl
489f1d8: Include more of sys/types.h and sys/signal.h under NaCl
02b55a2: Define _POSIX_THREADS under NaCl.
9f95ad0: Fix computed jump calculations in ARM memcpy

TEST= toolchain bots + toolchain_build/toolchain_build.py
BUG= https://code.google.com/p/chromium/issues/detail?id=338220
R=mcgrathr@chromium.org

Review URL: https://codereview.chromium.org/146703007
------------------------------------------------------------------------

Sign in to add a comment